Friday, May 20, 2011

Explicit versus Implicit Instruction in Phonemic Awareness

 Cunningham, Anne. (1990) Journal of Experimental Child Psychology, 50. 429-444.

Early (pre-NRP report) study of phonemic awareness instruction, using two forms: 1) a metalevel approach (emphasized applying knowledge, the the utility of PA knowledge, and its value in relation to reading) and 2) skill and drill - procedure of segmenting and blending phonemes. 42 K and 42 first-graders in the 10 week training study. Results supported causal relationship between PA and reading achievement. Both groups showed significant improvement in reading, however, the meta-level group performed significantly better on transfer tasks (untaught correspondences). [Control group listened to a story, answered questions, and discussed story.]

-Achievement test of reading ability (Metro Rdg. Achievement Test L1 & Primer)
-Aptitude (Otis-Lennon)
-Phonemic Awareness  (selected from previous studies - including Bradley & Bryant, and Stanovich et al '84)
     -phoneme deletion task ( 3 practice and 10 test items)
     -phoneme discrimination task (Lindamood Auditory Conceptualization test)
         Discriminating sounds, perceiving how many and in what order in spoken words (used colored blocks.)
     -phonological oddity rhyme task (Bradley & Bryant, '78, '83)
         Initial, medial, final sounds in words compared.

Results and Discussion
     The level of PA for both grades and treatments as the same except the K meta-level group were statistically better on phoneme deletion. First grade reading achievement improved significantly as the children were receiving reading instruction in the classroom and were much better at applying PA knowledge on reading tasks than the skill and drill group.  Hierarchical multiple regression analyses with PA tests and IQ combined as the predictor variables with the Metro scores as the criterion variable.  PA measures accounted for 60% of the variance in K and 51% in first grade reading in the spring.

     Results indicate that training in PA improves children's reading. The hypothesis that PA is a consequence of learning to read was rejected since training had a significant effect on reading achievement.  Both training groups performed significantly better on PA tasks and reading than the control group who had ordinary reading instruction in their respective grades.
     The metalevel instruction that emphasized "interrelations between phonemic awareness and the process of reading, motivation to use phonemic awareness in decoding, and specific strategic behaviors to implement phonemic awareness was a more effective program of instruction than a skill and drill approach that taught the component skill in isolation." Explicit instruction in segmenting and blending helped children transfer and apply PA skills to reading. While both groups learned PA, only one (metalevel) could effectively transfer their knowledge to the reading task. Skill & drill may teach the component skills but children don't make connections to apply them without a metacognitive rationale for them.


No comments:

Post a Comment